10/12/15  

1  

Portable Sleep Testing in Hospitalized Patients Rami Khayat, MD

1

Heart Failure AND Public Health •  ≈ 6 million Americans with heart failure (>2%

population •  20 million people with asymptomatic cardiac

impairment •  400,000 - 700,000 new cases diagnosed each

year •  Most frequent cause of hospitalization in patients

older than 65 years •  > 1.1 million heart failure hospitalizations annually

2 Redfield et al circulation 1998 And AHA Fact book

Trends in Length of Stay and Short-term Outcomes Among Medicare Patients Hospitalized for Heart Failure, 1993-2006JAMA. 2010;303(21):2141-2147. doi:10.1001/jama.2010.748

3

10/12/15  

2  

•  Heart failure causes or contributes to 250,000 deaths/year

•  1-Year mortality rate is about 10-15% •  5-Year mortality rate approaches 50% •  25% readmission rate at 1 month; 50% at 6 months •  > $18 billion in annual direct costs •  Despite current therapies and disease management

approaches, the rate of heart failure hospitalization and mortality remains unacceptably high

Impact of Heart Failure on Public Health

4

Impact of Heart Failure on Public Health

§ 83% of all heart failure patients will be hospitalized once and 43% at least 4 times.

§  In 2010, there were 1.023 million admissions with HF in the US, essentially unchanged from 2001.

§ By 2030, the total cost of HF will increase by 120% to $ 70 billion from 2013 estimated cost of $32 billion.

§  AHA Heart Disease and Stroke Statistics 2013

5

Roger et al Circulation 2011 Oldenburg et al Eur J Heart Fail 2007

Sleep Disordered Breathing (SDB)

Obstructive Sleep Apnea (OSA) Central Sleep Apnea (CSA)

Cheyne-Stokes Respiration Idiopathic Central Sleep

Apnea Syndrome

“Mixed SDB”: PAP induced CSA Mixed CSA-OSA (heart failure)

6

10/12/15  

3  

PR 56 y/o with dilated cardiomyopathy: inpatient sleep study at end of decompnesation episode

JS: Ambulatory 58M with Afib and snoring

10/12/15  

4  

Patient AH, 79 y/f studied on O2

LP 45 y/o male with ADHF

63 y/o OSA with long circulatory time

10/12/15  

5  

Cheyne Stokes Respiration

ADHF patients hospitalized between 12/30/06 and 1/31/08 (n=559)

ADHF and no known SDB (n=466)

Unsuccessful recording (n=71)

ADHF patients with successful recording (n=395)

No SDB (n=97) 25% SDB (n=298) 75%

OSA (n=226) 57%

OSA validation cohort) (n=111)

Patients returned (n=62) 56%, all

classified as OSA

CSA (n=72) 18%

CSA validation cohort) (n=26)

Patients returned (n=12) reclassified

as OSA

Excluded patients with ADHF and self reported existing SDB (n=93)

Khayat et al JCF-2009

OSA Mean (SE)

CSA Mean (SE)

Negative Mean (SE)

Negatives vs. OSA P value

Negative vs. CSA P value

CSA vs. OSA P value

Age 60 (0.9) 58 (1.8) 56 (1.6) 0.03 0.37 0.40

Male 69% (3%) 75% (5%) 38% (5%) 0.0001 0.0001 0.30

Ischemic

Dilated

Others

62% (3%)

23% (3%)

15% (2%)

64% (6%)

14% (4%)

22% (5%)

44% (5%)

35% (5%)

21% (4%)

0.003

0.02

0.22

0.01

0.001

0.80

0.82

0.11

0.16

LVEF 34 (1.2) 27 (1.7) 38 (1.8) 0.06 0.0001 0.0008

BMI 33 (0.6) 29 (0.9) 31 (0.8) 0.03 0.12 0.0001

LVEDD 57 (1.1) 63 (1.6) 54 (1.2) 0.14 0.0001 0.0037

BNP 746 (66) 1341 (161) 873 (130) 0.35 0.02 0.001

A-fib 39% (3%) 32% (6%) 28% (5%) 0.06 0.57 0.31

Khayat et al JCF-2009

Characteristics of ADHF Patients by SDB Status

15

10/12/15  

6  

Variable Names Pearson Correlation Coefficients

Number of Patients

95% Confidence Limits

AHI with EF -0.10 370 (-0.20, 0.00) AHI with LVEDD 0.19 281 (0.07, 0.30) AHI with BMI 0.17 393 (0.07, 0.26) AHI with A-Fib -0.02 394 (-0.12, 0.08) AHI with Age 0.02 395 (-0.08, 0.12) AHI with BNP 0.004 294 (-0.11, 0.12)

Predictors of SDB in all patients with ADHF

Variable Names Pearson Correlation Coefficients

Number of Patients

95% Confidence Limits

AHI with EF -0.01 279 (-0.13, 0.11) AHI with LVEDD 0.15 212 (0.01, 0.28) AHI with BMI 0.18 298 (0.07, 0.29) AHI with Age -0.06 298 (-0.17, 0.05) AHI with BNP -0.006 231 (-0.13, 0.12)

Predictors of AHI in all patients with ADHF and SDB

Variable Names Pearson Correlation Coefficients

Number of Patients

95% Confidence Limits

AHI with EF 0.07 212 (-0.07, 0.20) AHI with LVEDD 0.11 161 (-0.05, 0.26) AHI with BMI 0.30 226 (0.17, 0.41) AHI with A-Fib -0.09 225 (-0.22, 0.04) AHI with Age -0.01 226 (-0.14, 0.12) AHI with BNP -0.07 174 (-0.22, 0.08)

Predictors of severity of OSA

Variable Names Pearson Correlation Coefficients

Number of Patients

95% Confidence Limits

AHI with EF -0.08 67 (-0.32, 0.16) AHI with LVEDD 0.09 51 (-0.19, 0.36) AHI with BMI 0.39 72 (-0.19, 0.27) AHI with A-Fib -0.05 72 (-0.28, 0.19) AHI with Age -0.16 72 (-0.38, 0.07) AHI with BNP -0.05 57 (-0.31, 0.21)

Predictors of severity of CSA

10/12/15  

7  

Comparison of AHI between the in-hospital study and the polysomnography in the validated OSA and CSA patients  

   

PSG AHI   Inpatient AHI  Difference between

PSG AHI and Inpatient AHI  

95% CI for the Difference  Mean (SE) (N)   Mean (SE) (N)  

OSA  41.7 (3.9) (62)   37.4 (2.5) (62)  

4.3   -1.1, 9.6  

CSA  36.4 (7.2) (12)   49.1(5.9) (12)  

-12.7   -29.9, 4.5  

Persistence of SDB and validation of the Inpatient Testing

Khayat et al JCF-2009

20

Comparison of Distribution of Cardiac readmissions in 6 months between CSA and no SDB Patients

Note the higher percent of patients readmitted for each count in the CSA

Khayat et al JCF 2012

10/12/15  

8  

Negative binomial models for count of readmissions within 6 months- ADHF patients with HFrEF

Model  

Rate Ratio (95% Confidence Interval) p-value  

CSA vs Negative   CSA vs OSA  

OSA vs Negative  

Univariate   2.0 (1.6, 2.5) p < 0.0001  

1.3 (1.0, 1.5) p = 0.02  

1.6 (1.3, 2.0) p < 0.0001  

Adjusted   2.0 (1.5, 2.6) p < 0.0001  

1.2 (1.0, 1.5) p = 0.05  

1.6 (1.3, 2.1) p < 0.0001  

23

SDB is Independently Associated with Post-discharge Mortality 24

Khayat et al EHJ 2015

10/12/15  

9  

Treatment of SDB during ADHF and in the Post-Discharge Period

25

Baseline Characteristics Control group (n=23)

Mean (SE)

Intervention group (n=23)

Mean (SE)

95% CI for the difference

LVEF (%) 25.4 (1.8) 26.3 (1.8) (-6.1, 4.3) AHI (events per hour) 33 (3.0) 36 (2.0) (-11, 5) BNP 1154 (261) 1117 (259) (-708, 780) SBP (mmHg) 110 (4.0) 107 (4.0) (-8, 14) DBP (mmHg) 67 (2.0) 66 (2.0) (-5, 8) Heart Rate 88 (4.0) 81 (3.0) (-3, 17) Creatinine 1.3 (0.08) 1.4 (0.13) (-0.4, 0.2) LVEDD 62 (2.0) 64 (2.0) (-8, 2) LVEDV 227(18) 235 (20) (-63, 47) LVESV 169 (15) 171 (17) (-48, 44) Age 58 (3.0) 55 (3.0) (-5, 11) BMI (Kg/m2) 32 (2.0) 35 (3.0) (-10, 3) % Male 83 (8.0) 65 (10) (-9, 44) % Ischemic cardiomyopathy 78 (9.0) 87 (7.0) (-32, 14)

% of b-blokers 83 (8.0) 74 (9.0) (-16, 34) % of Number ACEI or AII-I 65 (10) 48(11) (-12, 47) % Diabetes 54 (11) 61 (10) (-41, 20)

Pilot RCT Evaluating Treatment of OSA during ADHF; Khayat et al Chest 2009

Effect of In Hospital APAP on Cardiac Function three days post-randomization

Control Treatment Difference (APAP Effects)

(p-value) LVEF

three days post randomization 25.8 30.4 4.6

(0.031) change from base line -0.2 4.4

(Final – Baseline) LVESV

three days post randomization 169 144 -24.8

(0.0007) change from base line 3.2 -22.1

(Final – Baseline) LVEDV

three days post randomization 228 204 -23.9

(0.03) change from base line 2.1 -22

(Final – Baseline)

Khayat et al Chest 2009

10/12/15  

10  

Effect of PAP on hemodynamic and neurohumoral measures

Change from baseline Control Treatment Difference (p-value)

Urinary Norepinephrine -0.003 -0.011 -0.0008 (0.18)

BNP 17 -457 -474

(0.13)

BUN -2.4 1.6 4.1

(0.18)

Creatinine -0.03 0.21 -0.24 (0.19)

Systolic BP -5.2 -6.5 -1.3

(0.78)

Diastolic BP -1.2 -3 -1.9 (0.6)

Weight -0.5 -2 -1.5

(0.048)

Khayat et al Chest 2009

Pilot RCT Evaluating the Effect of In-hospital PAP on Discharge LVEF in Patients with ADHF and OSA

3 days post randomizationBaseline

50

40

30

20

10

3 days post randomizationBaseline

APAP Group

LVEF

Control Group

Khayat et al Chest 2009

29